Research On Dynamic Characteristics And Fault Diagnosis Of Deep Groove Ball Bearing

As an important part of mechanical system,bearing dynamic characteristics analysis and fault diagnosis research is of great significance.In this paper,the deep groove ball bearing at the fan end of the motor is taken as the research object,and the simulation parameters are set according to the actual working conditions.Considering the influence of Hertz contact and elastohydrodynamic lubrication,the equivalent stiffness calculation model of ball and raceway contact was established.Combined with relevant theories of bearing contact,equivalent stiffness and friction,the dynamic characteristics of bearings were analyzed and the fault classification methods were studied as follows:1.Solidworks software was used to establish the three-dimensional model of bearings.Since the equivalent stiffness between the bearing ball and the inner and outer rings affects the dynamic characteristics of the rotating system,the equivalent stiffness calculation model was established considering the influence of Hertz contact stiffness and lubricating contact stiffness,and the influence of radial load,rotational speed and oil film thickness on the equivalent stiffness was analyzed.The results show that the equivalent stiffness between ball and raceway increases with the increase of radial load,and decreases with the increase of rotational speed and oil film thickness.2.Considering the effect of equivalent stiffness,the bearing pure rigid body model was established based on the rigid body theory,and the theoretical values of bearing kinematic parameters were compared with the simulation values to verify the correctness of the modeling method.The difference of contact force between ball and inner and outer ring under lubricated friction and dry friction is compared and analyzed.The contact force between ball and inner ring is simulated under two conditions of starting and stopping.Compare the contact force between ball and different components under the same working condition.The angular acceleration,linear velocity,contact force and cage angular velocity of ball were studied by setting different loads and rotational speeds.The rigidflexible coupling model of bearing was established based on the rigid-flexible coupling theory.In the modeling process,the influence of flexibility and equivalent stiffness of the outer ring was considered.Under the same working condition,the rigid-flexible coupling model and the pure rigid body model were compared.3.Simulation data and actual data were taken as input signals and processed by empirical mode decomposition(EMD).Hilbert envelope processing was performed on the decomposed intrinsic mode component(IMF)with the highest correlation.Finally,the first three peak values of the envelope processing fault frequency were selected as fault samples,and the Sparrows algorithm(SSA)improved support vector machine(SVM)and comparison algorithm were used to classify the faults.The classification results show that selecting the fault frequency as the input feature vector can accurately classify the fault category in the support vector machine improved by Sparrow algorithm.This method provides a high accuracy method for bearing fault classification.